Composição, distribuição e organização espacial das lianas em uma floresta estacional semidecidual em São Carlos-SP.

Abstract

Many authors did increasingly study lianas in the last years. This group of plants is scarcely known from the point of view of biology, and mainly ecology and their interactions in forests. Great interest remains on aspects related to forest fragment management. Therefore, the main goal of this study was to continue the preliminary survey on occurrence of lianas in a semidecidual tropical forest fragment, in São Carlos-SP, Brazil, measuring occurrence, distribution and organization. At the forest border and inner sites ten 10 x 50 m plots were established, where liana with a diameter equal or greater then 1 cm, and trees with a diameter equal or greater then 5 cm, at 1,3 m from ground surface, were identified and measured. Hight and crown diameter of trees were measured, and the host and the host parts where liana did settle were identified. As hosts were considered trees with a diameter at breast height (DBH)equal or greater then 5 cm, trees smaller then 5 cm at DBH, and lianas used as stand. Settle points were: crown, liana and more then one settle point. Also the partaking of liana on forest litter composition were studied. For this survey, 23 collectors with 50 x 50 cm hanging 50 cm above ground surface were used. As result, 60 species with 2,150 lianas, and 476 trees were found. In general, number of liana at the border were greater than at the inner sites, being 901 on border crown and 733 on inner crown. Tree diameter and hight were significantly correlated with number of liana at the border (r = 0.81; P = 0.09), but not at the inner sites (r = -0.71; P = 0.17). Crown size was not significantly correlated with number of liana at the border and inner plots (respectively r = -0.30; P = 0.68 and r = -0.80; P = 0.13), but with no different between the sites. The aggregation index did show an aggregated distribution for trees at the border sites, and randomized at the inner plots. For lianas, this index points to an aggregated distribution, many times favored by previous liana settling the host. Evaluation of the structural descriptors like number of liana, number of trees, number of trees with liana, mean stem diameter, tree hight and crown size, using the main component analysis, did not find differences between border and inner plots. Although the correspondence analysis for liana related to the site, did point to some species more specific to border or to inner plot habitats. Litter measurements, each 15-days, yielded up to 12.220 kg ha-1 year-1. Leaves did contribute with 7,750 kg ha-1 year-1, 18.96% being liana leaves. Litter yield of liana leaves was related to the period of flowering and fruiting from their main species. Although there was no clear standard for leaf litter fall. Further studies are needed on the relation with tree species, to allow better management mainly of semidecidual forest fragments.